Message266314
| Author | vstinner |
|---|---|
| Recipients | Adrian Wielgosik, Demur Rumed, mark.dickinson, serhiy.storchaka, vstinner, ztane |
| Date | 2016-05-25.06:58:51 |
| SpamBayes Score | -1.0 |
| Marked as misclassified | Yes |
| Message-id | <1464159531.85.0.506897340622.issue27097@psf.upfronthosting.co.za> |
| In-reply-to |
| Content | |
|---|---|
Antti Haapala: > However what I'd prefer here as the size of the type is important, to use `uint16_t` or typedef instead of just `unsigned short`, which is a statemeent that the value must be at least "2 bytes wide". I'm not aware of a platform where sizeof(unsigned short) is not *exactly* 16-bits. At least, no platform supported by CPython. Include/unicodeobject.h is quite explicit: #if SIZEOF_SHORT == 2 typedef unsigned short Py_UCS2; #else #error "Could not find a proper typedef for Py_UCS2" #endif Maybe we should use a similar code in ceval.c? #if SIZEOF_SHORT != 2 # error "Could not find a proper type for 16-bit pointer" #endif |
|
| History | |||
|---|---|---|---|
| Date | User | Action | Args |
| 2016-05-25 06:58:51 | vstinner | set | recipients: + vstinner, mark.dickinson, serhiy.storchaka, ztane, Demur Rumed, Adrian Wielgosik |
| 2016-05-25 06:58:51 | vstinner | set | messageid: <1464159531.85.0.506897340622.issue27097@psf.upfronthosting.co.za> |
| 2016-05-25 06:58:51 | vstinner | link | issue27097 messages |
| 2016-05-25 06:58:51 | vstinner | create | |